A risk score isn't a target. It's a description.
The risk adjustment factor score is how Medicare measures how sick a group of members really is. It shapes how care and support are planned for them next year. When it's accurate, the plan's resources line up with the people who need them most. That is the whole case for risk adjustment accuracy, and it is a case about patients.
But somewhere along the way, a lot of organizations started treating the score like a number to push upward. Reviews in the fourth quarter. Chart searches after the fact. Lists of diagnoses to "capture" before the year closes. That gets the order backward, and it changes what a risk team spends its days doing.
How does a description turn into a target?
Nobody sets out to distort a record. The drift is quieter than that. The score is visible, it moves, and it can be compared from one year to the next. Anything that behaves like that ends up on a slide, and anything on a slide ends up with a goal attached.
Once there is a goal, the work reorganizes around it. Success becomes the conditions added this year. Suspected conditions turn into a list of things to capture, rather than questions to answer. A suspect that a clinician rules out shows up in the report as a missed opportunity. Leaders do not have to ask for any of this. It follows from what gets counted.
It is the same trap we described in measuring care management by calls made: the number that is easiest to count, and that goes up when people push harder, slowly becomes the job.
The score should go up when the record finally matches how sick a member is, and go down when a condition was never really there. Both are wins.
Why does ruling out a condition count as much as confirming one?
Because the goal is a true record, and a true record has two ways of being wrong. It can leave out a condition the member really has. Or it can carry a condition the member does not have, or no longer has.
Picture a composite visit, as an illustration. A primary care doctor sees two suspected conditions on the summary for the next patient. The first is supported by an earlier clinician's note and a trend across recent lab results. The doctor asks about it, examines the patient, and confirms it. The second rests on a medication in the patient's history. When asked, the patient explains it was prescribed for something else entirely. The doctor rules it out.
In a lift-driven program, only the first decision counts. The second looks like a suspect that got away. But both made the record more accurate, and both took the same clinical attention. The second may matter more than it looks: a condition that sits in a record without being real shapes care plans, outreach, and support for a patient who doesn't need them.
There is a practical reason too. Clinicians notice what a suspect list is for. If the list only ever wants a yes, they learn to treat it as paperwork. If ruling out is treated as a finished, valued answer, the list becomes what it should be: a set of honest clinical questions, each with its evidence attached.
What does chasing the score cost patients and care teams?
Look at the work a lift-driven year produces. Reviews pile into the fourth quarter, after most visits are over. Teams search charts for diagnoses supported somewhere in the record but never addressed with the patient. Clinicians get queries about visits they barely remember. The record fills in, but by people who were not in the room.
- For patients, a condition worth discussing was not discussed when they were sitting in front of their doctor. If it was real, the conversation never happened. If it was not, it may end up in the record anyway.
- For clinicians, the work arrives as queries and reminders long after the visit, stacked on top of the patients in front of them now.
- For the organization, diagnoses added through after-the-fact chart review are the kind that draw federal audit scrutiny. We've written about why accuracy and audit exposure pull apart when the work happens late.
- For planning, a record that is wrong in either direction sends care management, outreach, and support to the wrong members.
None of this means the gaps aren't real. Chronic conditions have to be documented every calendar year to count, and many are not. When they are missed, the population looks healthier on paper than it is, and the resources available to care for those members shrink with it. The question is not whether to close that gap. It's whether the program is built to describe members accurately, or to move a number.
What does a program built for risk adjustment accuracy look like?
The difference is less about effort than about where the work happens and what counts as done. ElevateCare 360's Risk Module is built around accuracy, not lift:
- Before the visitSuspected conditions surface ahead of the appointment, with the evidence behind each one drawn from claims history, medications, lab results, and prior documentation, and linked to its source.
- In the roomThe clinician confirms or rules out each one with the patient in the room. Ruling out counts as much as confirming. The clinician makes the call; the suggestion only informs it.
- At the point of careDocumentation happens during the visit, not in a chart review months later. Only what the clinical evidence supports is documented, so the record reflects what was actually evaluated with the patient present.
This is the timing argument we made in why prospective risk adjustment is about when the insight arrives. The point here is the goal underneath it. Moving the work into the visit only helps if the visit is allowed to say no.
What changes in a risk team's week?
- The question in the review meeting changes from how far the score moved to whether this year's record is true, and where it still isn't.
- A suspected condition is finished when a clinician has evaluated it, confirmed or ruled out, not when it has been captured.
- Clinical documentation staff spend their time preparing upcoming visits instead of reopening finished ones.
- Leaders see where documentation gaps cluster by provider, panel, or condition, while there is still time in the year to support those clinicians.
- The year-end chart review becomes a backstop rather than the plan.
Describe your members accurately, and the care plan follows. For how the pre-visit view works, see ElevateCare 360's approach to risk adjustment accuracy.
Frequently asked questions
- What is risk adjustment accuracy?
- It means the documented conditions in each member's record match how sick that member actually is: nothing real left out, and nothing included that isn't there. The risk score is a summary of that record, so an accurate record produces a score that describes the population rather than one that has been pushed in either direction.
- Is it a problem if a member's risk score goes down?
- Not if it went down because the record became more accurate. A score that drops after a clinician rules out a condition that was never really present is describing the member better than it did before. The goal is a true record, not a higher number.
- Doesn't ruling out conditions mean fewer resources for patients?
- Resources are meant to follow real need. A condition that isn't present directs care management and outreach toward a patient who doesn't need it. Ruling it out, and confirming the conditions that are real, keeps resources aligned with the members who need them most.
- How should a risk adjustment program measure success?
- By whether suspected conditions get evaluated with the patient present and documented accurately, confirmed or ruled out, and by where documentation gaps still cluster. Measuring only how far the score moved rewards adding conditions and treats every ruled-out suspect as a loss.